YOUR employer could soon use your smartwatch to check how hard you are working. An artificially intelligent algorithm predicts whether someone has high or low workplace performance based on data from their electronic devices. However, there are concerns about how such an algorithm could be used.
Andrew Campbell at Dartmouth College in New Hampshire and colleagues compared 554 people鈥檚 performance at work with data from their smartphones, fitness-monitoring wristbands and location trackers. They monitored when people were at their desk, as well as data on sleep, heart rate and estimated stress levels.
The participants, most of whom worked at either a tech company or a financial services consultancy, had to assess their own productivity each week by answering questions about tasks they had completed or neglected, and whether they tried to benefit their team or organisation.
Advertisement
Based on people鈥檚 answers, Campbell鈥檚 team classified them as either 鈥渉igher鈥� or 鈥渓ower鈥� performing. Given a week鈥檚 worth of tracking data, the AI could determine that classification with 83 per cent accuracy (Proceedings of the ACM, ).
Campbell believes tracking data could be used to provide objective feedback to employees. He says the technology could eventually replace subjective reviews of workplace performance.
However, the algorithm is still very crude. It only puts people into one of two categories and its accuracy is still quite low. Additionally, it was trained on subjective participant surveys, which aren鈥檛 always accurate.
There is a risk that the tech could be used uncritically to judge someone鈥檚 work performance. Previous studies have found that people are more willing to trust an opinion if they think it comes from an algorithm, even if there is no good reason to do so.
Ensuring that workers have a say in what information is tracked and have the chance to contest data if the need arises is critical, says Naja Holten M酶ller at the University of Copenhagen in Denmark.
鈥淭he technology could be used to eventually replace subjective reviews of workplace performance鈥�
Some companies are racing ahead with similar technology. The gaming firm Activision Blizzard pays employees to track their fertility, mental health and diet, reportedly using aggregated data to help lower medical insurance costs, for example.
Even though projects like this are often voluntary, there can be a social pressure to participate, says Holten M酶ller.
Campbell admits there are privacy and ethical concerns. 鈥淭he worker would have to be in control of access to their data and give consent to organisations,鈥� he says.
But he doesn鈥檛 rule out approaching companies to use the algorithm on their workers in future.
